MS-SBVAR: change default for drop option

matlab/ms-sbvar/initialize_ms_sbvar_options.m

@@ -87,8 +87,8 @@ options_.ms.random_function_convergence_criterion = 0.1;
 options_.ms.random_parameter_convergence_criterion = 0.1;
 % simulation
 options_.ms.mh_replic = 10000; % default differs from Dan's code
-options_.ms.drop = 0.1*options_.ms.mh_replic;
 options_.ms.thinning_factor = 1;
+options_.ms.drop = 0.1*options_.ms.mh_replic*options_.ms.thinning_factor;
 options_.ms.adaptive_mh_draws = 30000;
 % mdd
 options_.ms.mdd_proposal_draws = 100000;

preprocessor/ComputingTasks.cc

@@ -966,10 +966,11 @@ MSSBVARSimulationStatement::writeOutput(ostream &output, const string &basename)
 
   // Redeclare drop option if necessary
   OptionsList::num_options_t::const_iterator mh_replic_it = options_list.num_options.find("ms.mh_replic");
+  OptionsList::num_options_t::const_iterator thinning_factor_it = options_list.num_options.find("ms.thinning_factor");
   OptionsList::num_options_t::const_iterator drop_it = options_list.num_options.find("ms.drop");
-  if (mh_replic_it != options_list.num_options.end())
+  if (mh_replic_it != options_list.num_options.end() || thinning_factor_it != options_list.num_options.end())
     if (drop_it == options_list.num_options.end())
-      output << "options_.ms.drop = 0.1*options_.ms.mh_replic;" << endl;
+      output << "options_.ms.drop = 0.1*options_.ms.mh_replic*options_.ms.thinning_factor;" << endl;
 
   output << "[options_, oo_] = ms_simulation(M_, options_, oo_);" << endl;
 }